redis
Redis operations: data structures, expiry, persistence, Lua/transactions, clustering, cache patterns. Use for caching, queues, rate limiting, sessions; research version docs first.
Files1
View on GitHubSKILL.mdskills/databases/redis/SKILL.md
GitHubredis
Memory is fast and forgetful: design expiry, eviction, and persistence before trusting it.
Activate when
- Adding caching, distributed locks, rate limiters, sessions, or lightweight queues on Redis.
Do NOT activate for
- Durable system-of-record storage (see fundamentals, postgres-operations).
Procedure
- Version: detect server version; research in https://redis.io/docs/ for that version (commands and clustering behavior evolve).
- Structures: pick by access (strings, hashes, sorted sets for rankings/queues, streams for logs/queues, HyperLogLog/bitmaps for analytics); set TTLs deliberately; eviction policy matched to workload.
- Correctness: Lua scripts or transactions for check-and-set; Redlock-style locks only with fencing for truly exclusive work; idempotent consumers for stream/queue use.
- Persistence/HA: RDB+AOF per durability needs; replicas + sentinel/cluster understood before promising failover.
- Security: AUTH/ACLs, no public exposure, sensitive values encrypted client-side where required.
- Verify: eviction/expiry behavior tested, failover drill, memory-growth monitored, lock contention measured.
References
../fundamentals/,../database-security/.- Registry:
../../research/documentation-search/references/framework-registry.yaml(redis).